Photography

Photography for evaluations can be considered an extension of, or variation on, participant observation [Page 316]or document analysis. Either existing or new photos can be used for evaluation purposes and can be analyzed either qualitatively or quantitatively.

It is interesting that photos are an underused form of data for evaluation purposes and that little has been written about them in relation to evaluation practice. Exceptions include Templin (1981) and Hurworth and Sweeney (1996).

Photography is particularly useful in evaluations in which

• there is a range of activities to be documented
• program participants are unable to undertake other forms of data collection (such as surveys or in-depth interviews). Examples include small children, non-English speakers, the physically and mentally handicapped, and the general public
• the visual image would be more “hard hitting” than prose
• programs are highly visual in themselves (e.g., programs in art, dance, drama)
• a program's effects lead to marked changes over time, such as before and after a land-care program, where an initial picture shows an eroded slope and the same slope, postintervention, is shown to have been reforested
• the physical and locational context of a program is important (e.g., for postoccupancy evaluations carried out by architects)

Photographs can add a great deal to evaluations. They provide visual communication that helps to develop insights, imagery not available by other means, the means to evoke powerful emotional responses, improved understandings about program contexts, assistance in documenting cultures, and a way of creating variety within reporting.

However, selection and analysis of photographs have to be undertaken rigorously. So, photos must not be just “happy snaps” but be sampled properly. Sampling techniques advocated by Templin include the following:

• Fixed-time sampling. Pictures are taken from a fixed position at regular intervals.
• Sampling across time. In this case, the photographer moves around, taking regular photos of a full range of activities.
• Event-based sampling. Only events in a certain category are photographed to address particular evaluation questions.
• Dimensionally based sampling. This is desirable when constraints of time make it impossible to cover all events. In this case, in conjunction with an evaluation question, a possible range is chosen.
• Shadow sampling. One person or group is followed throughout a certain time period or routine.
• Snowball sampling. A stakeholder is asked what, or who else, might be photographed.
• Theoretically informed sampling. Photographs are taken from different vantage points. For instance, one might take photos nearer the ground to see the world from a child's perspective or lie down on a hospital bed to see a ward from a patient's viewpoint.

Aside from rigorous sampling, there are other ways to enhance trustworthiness. These include the following:

• Ensuring there are no alterations or distortions so that photographs are evaluation evidence rather than “art”
• Ensuring that the photo is characteristic of what it purports to show
• Using photos as part of multimethod triangulation
• Explaining any selection bias caused by the client, the political situation, or social taboos
• Determining and explaining any influence the photographer might have had on the behaviors of those being photographed (prolonged engagement will help to avoid this)
• Posing a strong evaluation question so that there is a reason for the image and its content

Finally, ethical issues associated with the presentation of photographs in evaluation reports abound. If individuals and sites can be identified, informed consent will be necessary. This issue may be alleviated if photographs are taken from behind or from the side.

...